Tyme Group's Data and AI-Fueled Unicorn Journey

In this episode, Dietmar Böhmer, the Chief Analytics Officer of Tyme Group, discusses the company's strategy for integrating data and analytics teams within business units to maximize efficiency. He also delves into Tyme Group's AI Roadmap, highlighting innovative applications aimed at enhancing developer productivity and improving customer experiences.

Tyme Group, a digital banking company with its headquarters in Singapore, has experienced significant growth since its inception in 2019. The company achieved unicorn status in December 2024 after securing $250 million in a Series D funding round. Tyme Group has expanded its reach to serve 15 million customers across Africa and Southeast Asia through its digital banking subsidiaries, TymeBank in South Africa and GoTyme Bank in the Philippines.

Transcript

AIBP: 0:00

In the fast paced and fiercely competitive world of digital banking, success is built on a foundation of data driven insights and strategic innovation.

Dietmar - Tyme: 0:08

Being a digital player, data analytics is absolutely core to how we operate and what we do be extremely data driven and trying to optimize every piece of the business with facts and information.

AIBP: 0:23

Data isn't the end goal. It's a tool for clarity. To unlock its value, businesses must collect structure and interpret it effectively, reshaping how leaders address challenges and opportunities.

Dietmar - Tyme: 0:35

My definition of data is quite simple. It's information, and it's all about how do you use and structure that information to get the insights and answers to problems that you're trying to solve or also highlight problems that you might not not be aware of?

AIBP: 0:53

Data's potential is realized through a balance of innovation and regulation. Asia's measured approach towards artificial intelligence fosters innovation without stifling progress.

Dietmar - Tyme: 1:05

The other thing that they've not done, which you which you do see in other geographies, is that they've become too overbearing too quickly with regulation, where they try to regulate, AI too quickly, where, yes, obviously you need to have regulation. You can't just have people do whatever they want without without any guidance, but also you don't want to be too over, you don't want to push it too far, too quickly, to strive for innovation. So that's the thing in sort of the Asian countries, that really excites me about how they will embrace and honestly have the opportunity to leapfrog other geographies that have taken different different

AIBP: 1:48

The story of data in digital banking reflects a approaches. larger truth, success lies in leveraging insights to adapt, innovate and compete, driving progress in a constantly evolving landscape. Please enjoy this episode from the AIBP ASEAN B2B growth podcast.

Unknown: 2:10

The AIBP ASEAN B2B growth podcast is a series of fireside chats with business leaders in Southeast Asia focused on growth in the region. Topics discussed include business strategy, sales and marketing, enterprise, technology and innovation.

Vanessa Kwan: 2:36

Hello and welcome to the ASEAN B2B growth podcast, where we sit down with individuals responsible for driving growth within their organizations here in Southeast Asia. My name is Vanessa, and I'll be your host for today in this episode, we are excited to have with us. Mr. Ditma Bo, the chief analytics officer at time group, the Singapore based time group was crowned as Southeast Asia's newest tech unicorn just last month, time operates in South Africa through time bank and in the Philippines via go. Time Bank serving over 15 million users across both region our guest today built up and leads their analytics and data team. How exciting is that? But without a without further ado? Ditma, welcome and thank you for joining us, could I perhaps get you to give us a brief introduction of yourself?

Dietmar - Tyme: 3:25

Hi, Vanessa, thanks for having me. Yes, as you mentioned, did my BOMA the chief analytics officer of the time group. So my role and responsibility is to lead analytics AI and data capability for the time group. And yeah, I'm a passionate individual that just loves using data through the latest technology as well as machine learning and AI to solve complex problems as well as make a meaningful full impact. And yeah, I've got this amazing opportunity through through time to live my passion and and and drive financial inclusion, as we all strive for initially, when we when we set up the company.

Vanessa Kwan: 4:12

Thank you, ditna, for that introduction. You know, it's fascinating to hear about how you've come to where you are currently at time group. Perhaps I can also start off by going deeper into the organizational structure of the company, and you know how that influenced the strategy behind some of the data and AI initiatives that you have planned out. Can you walk us through your approach to structuring the data and AI teams and initiatives across the organizations? And you know, in the longer term, how does it kind of align with the overall digital banking strategy that you have?

Dietmar - Tyme: 4:47

Yeah, so that's a very important question, an important point being a digital player, data analytics is absolutely core to how we operate and what we do. We extremely data driven and. Trying to optimize every piece of the business with facts and information. So the way that we've gone about it is that we've set up our analytics capability as a group function. So we've got two or three group functions where we try and drive a group capability. But what we're not trying to do is we're not trying to set up a centralized analytics environment that just provides a service to stakeholders and staff. The way that we've set it up is that the analytics team, even though it's a centralized team, is embedded in the business. Because the only way, or at least my view, is that the only way that the analytics team can provide value to their stakeholders is to sit with their stakeholders and understand the business they need to be in the business. They need to be in that market, where they where they servicing, because if they don't have the context of the customer in that market, if they don't spend the day with the team that they supporting, then they don't really have the context of, what are the business problems that that the business unit that they supporting is is trying to solve, and how do you then provide proactive information and support for them to drive the business forward? So, so we've got a centralized team, but that is really just about driving Excellence, making sure people are using the best techniques, and this cross sharing across individuals, because people are still dedicated to a very specific area or stakeholder that they then service and have dedicated focus on a very specific area. So what we don't do is have a central team people log a ticket and say, We want to stand this better, or we want to build a specific model for a specific use case, and then it goes into some kind of prioritization backlog, or something like that. We've got very much dedicated capacity. They work very close with their stakeholders, because also having that trust relationship of them, being able to give independent advice of what they see in the in the business is also important, and that really just comes with with time, spending time with each other, building up that trust, having a relationship and yeah, that's the approach that we follow to to set up the the capability. So we've got an analytics team that's embedded into time bank in South Africa. Similarly, we've got a very similar team that's embedded into the Philippines business, servicing the stakeholders in that part of the business. But then also, we've got an analytics team sitting in our technology hub, which we run out of Vietnam. So we build and run all our banks out of our technology hub that's hosted in Vietnam. We've also got an analytics capability inside of that business that's servicing the product owners that sit there, as well as the different feature teams that they get insight in terms of the performance of those features, the funnels, the drop offs that they really understand, how customers are engaging with, the with with the products that they, that they building. We would, every once in a while, we would set up centralized capabilities that's more focused on R and D, for example. So like we've done 18 months ago, when the new wave of AI hit us, we set up a centralized capability to do a lot of research in early development in terms of, what is this technology? How can we use it? What are the different opportunities and but then from there on, we then disseminate that back into into into the businesses that they really unlock value using a lot of the research that was done potentially centrally to then unlock value in the in the different areas of the business

Vanessa Kwan: 8:43

Understand. Thank you very much for the very good overview of how the time group is kind of, kind of structuring and also leveraging data and AI for yourself and for the different subsidiaries, perhaps let's go back a little bit and explore some of like the fundamental concepts and practical applications of data and AI in the banking industry, you know you're a data professional. Could you perhaps give a very brief overview, in your words, what does data mean to you? And then also some examples, instances where data has helped, let's say, Time group, in generating actionable insights. You mentioned earlier, customers at the forefront of what you do as well? So perhaps something along those lines would be great.

Dietmar - Tyme: 9:25

Yeah. So my my definition of data is quite simple. It's it's information, and it's all about, how do you use and structure that information to get the insights and answers to problems that you're trying to solve, or also highlight problems that you might not not be aware of. If I if I can use maybe one example of something that I think is is a little bit unique to, I wouldn't say just our business model, but but digital banking in general. So originally, I had a long analytics career in traditional banks, and then eight and a half years ago, when I, when I moved over to time, we had to solve very different problems, and we had to think about customers very differently. If you're 100 year old organization and you've got a big base of customers and you've got a strong brand and trust, versus you starting up a new business and you, and you attracting and onboarding customers for the first time into your ecosystem, what you need to think about and what you need to worry about is very, very different. So for example, one of the big problems that we have to solve is driving customer adoption and understanding that adoption curve that customers go through, because when customers open a bank account through digital banks or means, it doesn't mean that these people don't currently have a way of living their financial life. It means they already have a bank account somewhere, or they found a way to live in the cash world or the informal financial sector, but they are in a habit of buying and selling and paying for goods in a very specific way, and now they're opening a new account, which is very seamless and great user experience. But they still have to change their behavior from what they're doing today to start using your your platform, so understanding what that looks like for different customers and how you're nudging along on those journey, what education is needed at what time is absolutely critical. So that's probably the one of the biggest problems that we had to solve and understand, and the only way that we could do that was with data as well as lots of experimentation to also generate the data that we that we needed. So that's probably the one example that I would highlight where we had to put very, very special attention, focus on, on to really understand what is that journey that customers go through, that we ultimately end up with as many as possible, highly active and entrenched customers on our platform as possible.

Vanessa Kwan: 12:06

Understand and more forward looking. Are there any newer or innovative case studies, you know, within the data and AI space, say, across the region or even globally, that you are looking at that you think can be applied to what the time group is currently doing? And then, you know, of course, time is one of many digital banks in this part of the world. For yourself and for other digital banks are like, how can you, you know, how can you prepare, essentially, for all these AI developments that are going by the hour?

Dietmar - Tyme: 12:36

Yeah, absolutely, as many, many people say, What a time to be alive if you, if you in this industry, it, it really is, is, is a time of lots of change, but also lots of unknowns. We obviously nobody knows exactly where this is going or where this is ending up, but I think we'll be starting to get a clearer and clearer view of what the what the capabilities are that one could extract and leverage from the latest wave of of AI. And once again, I think digital banks and digital players or tech companies are extremely well placed to take advantage of this, because they're already very much technology and data driven companies, where, historically 1010, years ago, a lot of you know, trying to get predictive insights or automation of process was very much a machine learning on AI problem, where, yes, you needed the data, needed quality data, but You needed quite sophisticated processes and techniques and individuals to build you those those models we're now AI has become a lot more commercialized, and it's actually gone back to becoming a data problem. So the better the quality of your data is, and the better you are exposing and making your data available into these AI ecosystems, the more value you can extract from it. And the big change that I foresee that is this going to come out over the next, let's say, 18 months, is that you're going to get a lot more autonomous AI agents that will perform tasks very similar to human staff members at the moment. So how I see this evolving is that we're going to end up with teams where where a manager was used to managing 10 individuals. He would still manage individuals, but he will also have five or six or seven virtual employees of these agents that are also performing very, very specific tasks, and the companies that are able to leverage and expose and get economies of scale through, through that are ultimately, I think, going to be the winners in this. So, yeah, how you think about not just structuring your data, but also. Do you think organisationally about structuring yourself and a manager that we've all learned over the last 100, couple of 100 years of how to manage people, how to incentivise people, how to get the best out of human beings, how to make sure that they fulfilled in what they do. But now, how do you how do you manage a virtual employee or an AI agent, how do you measure their contribution? How do you measure their quality of work? So this gonna, gonna be quite a, quite a fundamental change, I think, over the next 18 months, in terms of the the workforce, and especially companies that are already technology and data driven, that have the platforms to be able to do this? Yeah, I think are going to, definitely going to stop pushing the boundaries on some of these, these, these elements.

Vanessa Kwan: 15:50

Thank you for sharing. You know, it's very clear that both technology and data plays a critical role in a lot of modern business operations, not just within the Digital banks or even in the digital banking sphere. But, you know, let's kind of shift our focus a little bit to the more exciting world of AI. And we've also heard from your Vietnam team about some of the applications that they've been working on time access. Aipp, award winning submission incorporates AI across the entire operation. You know, to look at enhancing productivity, security, efficiency. But in the interest of time, perhaps we could focus on a couple that are a little bit more interesting exciting, namely the internal AI chat bot, as well as the AI driven threat detection to bolster cyber security. But before we get into that, could I maybe have you walk us through a little bit around times AI roadmap, and perhaps some insights into some of those initiatives.

Dietmar - Tyme: 16:50

Yeah, so if I sort of take a step back on AI and the way that we view AI and the applications of AI and how we focus on solutions that are AI driven, really see it across three different, different pillars. The first one is, is, is very much an efficiency and a quality play in our technology environment. And those are the ones that you've you've just mentioned, and that's in, how do we empower and get our developers to be more efficient? How do we get them to write code faster and better, but not just, not just that. How do we generate automated testing test case? How do we use AI to look at the code and come up with a lot of these test cases that we don't just have manual QA testers and UX testers? How do you start using AI to do that? And then that's where, for example, that the chat bot has become, come very, very important, because it's not just so, so it's something that we've trained up with, a lot of our principles of how we code and what are our standards. So it's very much aware of how we write our code, but we've also exposed our documentation of so. So we use Confluence as a centralized environment where everything gets documented. So every feature that gets gets gets built out, gets gets documented in a lot of detail, from the workflow to the final solution that's also embedded there. You can imagine that after eight and a half years of of building and literally making 30 to 50 production changes a day, that that Confluence environment is actually quite convoluted and big, and find something there becomes harder and harder. So putting your your sort of AI chat bot on top of that Google search to find information, but do it in a conversational way that you can find information about a feature that a different team might have built, or find out who that team is that has built this feature that you're trying to integrate with is proven to be hugely, hugely successful, as well as the bespoke coding assistant that we've that we've built, which is a plugging into the IDE that we use for the developers To write the code where it can help them refactor or optimize or debug code, because sometimes you on call, and at two in the morning you get phoned because there's a there's a there's a system issue, and you need to come and resolve the issue. It might not have been a piece of code that you've written. You might have taken it over for someone, or you just on call for the team, and it was written by a different team member. So just having this AI assistant to also just look at this code and help you interpret what is this code doing, has also proven hugely, hugely effective so that that's the first pillar, and that's really focusing on, how do we become more optimized and efficient in our our delivery and just getting product and features out faster. But then the other two features that we've also focused very much on is what we call AI for business, and that is just, how do we get our in country business operations also to be more efficient and effective? So whether or not it's automating a. Customer responses, or providing a proposed response to to a customer agent, or the automation of very repetitive automated task, which historically, you you wouldn't able to do a customer submit documents, and sometimes it's handwritten documents. Historically, OCR would not have been able to interpret that document, or find the pieces of information when I was LMS, they can read and they can identify who the customer is. It can create the case, and it can already automate a big part of the initial workflow to make your operational efficiency just so much higher that people, because nobody, enjoys opening a document trying to find who is this customer, and then looking up the customer. And if you can pre automate a lot of that, the you know, a lot of people talk about the threat and intimidation factor of AI, but if you actually build it in a way that that it takes a lot of the repetitive stuff that nobody's doing away, and people can really focus on on solving the problems, you also get, get a lot more, lot more buy in. So that's the second pillar. So is the AI for for business? And the last one is, is AI for customers? So how do we think about these AI features for for customers, whether or not that's in self service channels or dynamic features in our banking app, etc? So, so that's just a little bit of of insight in terms of how we think about we've got very dedicated focus. Because how you think about exposing an AI solution to a customer is not the same as how you would do that necessarily internally. So that's why we have those, those three, those three, three pillars and and hopefully in the first one, I gave you a bit of context of how we use that, in our in our technology, gi hub, with with the different solutions, and what that chat bot actually can do. So it's not just a general Q and A chat bot for internal staff, but it really is very much highly customized, using our code, using our documentation on all the features that we've that we've released into into production, just to make all of that information a lot more accessible and rich

Vanessa Kwan: 22:14

Understand. So to summarize is the three key pillars that you have, the first one to increase efficiency as well as optimization across all the initiatives. Second one, AI for business, and the third one, AI for customers. You know, on the flip side, we've heard a lot about, like, all the good things that AI can do for you. We've heard also a lot of challenges that organizations across Southeast Asia and I'm sure globally, face when it comes to, like, adopting different AI initiatives and projects across their organizations. You know for yourself in your experience, what would you say are some of the challenges that you face, that you think other people have also faced, and what would be your advice to them? I think one of the other key areas that we hear is a challenge is how a lot of the hackers attackers right now are using AI to reach in. So perhaps it'll be good to also have your opinion on that.

Dietmar - Tyme: 23:07

Yeah, so I'll focus on the first one, and then I'll also speak to a very, very specific problem that we had to solve, which you which you touched on when it comes to adoption. But absolutely, the reality is this technology is not just available for businesses and the good actors. Unfortunately, the bad actors have as much, if not more, access to the same technology. And in many environments, lots of people have, let's call it, a digital signature in the world, and whether or not it's your fingerprint or it's your face or it's your voice, lot of people use that for for authentication methods, whether or not it's at the point of onboarding, where you're showing an an ID, and they want to make sure that the individual and the ID is the same person. Or for repeat service, when you phone into into into a call center, making sure that you stay ahead of that curve, and you have the ability to determine an AI generated document, so an ID that was generated through through AI versus a picture that was generated through AI versus a voice, or even a video that's overlaid. Because even if you do a video call with somebody to try and verify that many companies have tried to do as a fallback, people can, these days, also just put filters on that you look like someone. So having these technologies run in the background to determine that is absolutely critical. And I'm not going to lie, it gets harder and harder, because this technology is maturing so fast and so quick, and if you don't have a dedicated focus on all these these aspects, where you could be exposed to that, it's a real challenge, because you do get these things at large scale where. Especially if you're an institution that's working with money, people are incentivized to try and get themselves into that, that ecosystem. And there's no ways that, as a human, that you can look at that if, when you were to look at the picture, the document, you will not pick up that it was whether or not it was government issued or or when it was ai, ai generated, and the only way that you can use it and detect that is through the use of AI that really goes deep into into these documents and the metadata of the of the information that's being being being submitted. So, so, so absolutely, that has to be a focus on anybody that has a customer facing or digital digital business. The second piece that I that I that I think is worth sharing, because it was a real challenge for us, initially in the business, because we all got very excited about AI and we saw lots of opportunities, but even with developing very customized solutions for business and individuals or teams, the adoption wasn't where we thought it would be, or people didn't embrace it necessarily at the rate that we thought they would, and we realized that we had to invest quite actively in the adoption of AI so so for every solution that we have, so as part of our AI team, we don't just have aI scientist and, let's say back in and front end engineers to package it as a, as a, as a end to end solution. We also have individuals that are AI adopters. So what they do is they actually sit with the team when we roll out a new solution into operations or into our technology environment, and then they actually sit and spend time for a week or two with the teams and go through them using that that we actually get feedback immediately in terms of what's working what's not working for that team. How can we improve the solution to to make it better, but also just to get people over this initial hurdle of start, start using it. Because the unfortunate reality is, Chat GPT is very good at at at many things. One of the things is actually providing an impressive but useless piece of information which gives people a false sense of what this technology can do, because everybody has played with ChatGPT and say, write me a dog about my company. Write me a poem about my dog, or write me a song about my company, and it does a great job, and everybody's got a big laugh. And I think by just giving this thing that one sentence, you are going to get amazing output. But you didn't ask it for something very specific, right? You asked it a very general question. And if you just write a general poem or song, it's fine, right? But if you wanted to perform a very specific task and get your very specific outcome, you have to be quite good, and have to refine and improve how you engage with this machine. And that's why we've invested in these, these AI adopters, to spend time with teams, to also improve their ability to engage with AI, to get the most out of out of AI. Because the reality is, it's a new skill that we all need to learn. For some people, it's more natural than others, but ultimately, we need to get people over that adoption curve to drive that so that's something that's given us immense, immense value by by investing in these individuals that that spend time on the ground to drive the adoption.

Vanessa Kwan: 28:37

That's something very interesting, and I'm hoping that a lot of other organizations out there have either considered what you're doing or are already doing what you're doing. It seems like a good way to actually get real feedback on the ground in terms of what's working, what's not working, and like you mentioned, the kind of details or information that you have to fit Chat GPT, for instance, to get the results that you want. So I think that's something certainly very exciting for us to stay tuned to in the next 510, years to see where that will bring us. You know, I think we've spoken a lot about data. We've spoken a bit about AI as well, but perhaps also looking at like the broader picture in terms of how technologies fall into time groups, larger business and innovation strategy. Could I also have you share a little bit more about what innovation mean to the time group, and how do you usually measure the success of those initiatives or efforts?

Dietmar - Tyme: 29:31

So there's, there's obviously different, different definitions of of innovation, but how we like to think and drive innovation is, how do you find creative ways or novel ways, or new ways to solve a problem? This could be an existing or an old problem, or it could be a new problem, one one of the things that we've done, which is actually come out of our dedicated. And ring fence innovation lab. So we've got somebody that heads up innovation for us was our kiosk and our kiosk model. So yes, we are digital bank, but we've got this, this, what we sometimes call our phygital model, which is a hybrid between digital and physical, but we've also got a physical kiosk where people can can onboard, but also not just open an account, but also within two, three minutes of spending the time to open an account, which in itself is very, very quick, you actually leave that kiosk with a card that's live and active. It's got your name, account number. It's a personalized card that you that would normally take, you know, a week to get posted by by most, most other banks. So that's an example of an existing problem that we that we gave to our innovation lab and said, How do we think about customer onboarding and card issuing in a different, different way, and and it's an example of something that that's that's proven very successful for us, because it helped us actually broaden the reach of our digital offering, where we don't just service people that are are completely digital, native and completely comfortable of only living in the in the in the app world, some People are still very nervous of that and hesitant of that, or some people just don't trust technology. And if they have a process to go through a kiosk where there's a human to assist them to and guide them through this process and get them up this digital curve, you know, it's always been one of our theories and beliefs, and it's something that that's proven itself to be quite successful, that even if it is successful, to onboard the younger, more innovative generation, how do we make sure that we don't leave anybody behind? Then we actually do have a proposition that does cater for the for the full for the full population and demographics

Vanessa Kwan: 32:00

Understand. So when you talk about like the innovation lab, where you have a separate leader who leads that part, you also encourage the younger generation to solve either existing problems or new potential problems that this thing might happen, hopefully not. So that's how you kind of go about solving the different challenges and different problems understand.

Dietmar - Tyme: 32:18

And I think the other thing that's that's also very important as part of that process. Just the last thing I want to add is, firstly, yes, it's important that you celebrate these successes. And we've got the weekly all hands, where the whole company comes together every Monday morning. And, you know, apart from just sharing information about where the company is, or big announcements, or people just giving updates from different teams, we also use that as a platform to recognize people that have solved problems in a creative way. But the other thing that I want to that that's actually the more important piece, is to also celebrate failures, because that's how you drive that culture of people not being scared of experimenting. And I'm not saying you you celebrate somebody that didn't think through a problem, and I just stumbled into something and it didn't work because they didn't apply their mind to it. But in a lot of cases, you actually learn a lot more from things that didn't work than things that did work. Because there's something that work. You actually don't do a post mortem, and sometimes you actually don't know you just got lucky, and you just go with it and and know that when you need to take it to the next step, you actually don't know what to do, versus when something didn't work, or there was a failure, and, and, and you don't blame a team that you actually celebrate it in a way where they share their learnings and insights, and How would they approach this problem the next time? I think it's also very, very important to drive an innovative culture that people are open and not scared to try new things.

Vanessa Kwan: 33:55

Understand and you know, going back to where we started off, this conversation was really looking at your role as the chief analytics officer for time group, and how you're currently structuring the data teams across the organization. What were some of the initial considerations you have when it comes to like, say, building, or, for example, buying, outsourcing different technology developments? You know, we understand also that that's TymeX that operates out of Vietnam. So are there certain situations or circumstances where you would consider one over the other?

Dietmar - Tyme: 34:30

Yeah, so that's a very good, good question, because the reality is, by partnering and outsourcing, in many instances, it's a great accelerator. But if you, there's also downside to if you, if you scale really quickly and you grow very fast, is that in many instances, you get to a point where you outgrow these, these, these Venus, that they can't keep up. You know, our business. That, you know, we launched in South Africa in 2019 and october 2022, we launched the bank in the Philippines, and we've got 15 million customers. So it scaled really quickly and and unfortunately, many of the vendors that we used in the early days did not, did not keep pace. So a lot of things we had to, had to bring in house, and anything that's critical to customer experience, so making sure that the app is up and running, making sure that this transaction performs in real time, all of those things we want to make sure that we've got the control and manage on that, that we're not dependent on a third party as much as much as we can. But there are instances where we've partnered with vendor that vendors that have just been absolutely fantastic. For example, we fully running on Cloud, we AWS, run and do everything in AWS, and we would not have been able to scale at the rate that we've done without having them as a partner and being in the cloud. And if we had to physically buy and procure physical infrastructure to keep up with with the pace of how our customers transactions grew, we would not have been able, able to keep up. So, so that's been a been a huge advantage for us, that sort of at the time that we launched was, was when, when cloud started maturing, and we could leverage and take advantage of that, and then also in the analytics space. We've we've had a great partnership with, with with Databricks. So we use Databricks as not just our analytics workbench, but we also use it for all our data engineering workloads we used as a data warehouse, and what that's really enabled us and empowered us to do is by using them as a as a partner, it takes away all the worries that we are to do and focus on the base layer of is the data available. Are the servers up and running is the data indexed? All of that just happens automatically, and all we have to do is we can worry about solving business problems. So we've got an engineering team that has to focus on being world class at ingesting data, because, as I mentioned, being a technology driven company, one of the things that our CTO is always very excited to share and tell people is that on good days, we release 30 to 50 new production releases per week, and sometimes we even do that on a day. We sounds amazing, but if you the data team, you can imagine that every change that you make to an IT system has an impact on the back end system, so the datas that we see either the databases or the information that goes into a database, so having the ability to adjust and compensate for that and still be able to consume that information and still provide the insights. So we, you know, we have, we had to build a world class capability of ingesting data, but then managing that infrastructure layer, we don't have to focus on, and then we can just focus them on, on solving business problems, driving insights, building models, serving the models, on, on, on that platform. So I think in in general, we've we've been burnt a lot by by by many of our technology partners, but they've been a handful that have, that have that have been, been been been absolute superstars in in helping us along this along this journey.

Vanessa Kwan: 38:31

It's good to hear that even though there were, there were some challenges that you faced, at least you managed to find partners like AWS, Databricks, who have managed to add value and help you scale as quickly as you did. You know, it's really good, really fascinating to hear about time's journey and approach to innovation to technology as well. And as we wrap up our conversation today, you know, I wanted to look a little bit more into the future and explore some of the broader industry trends. You know, in your opinion, what areas within AI are you personally excited number one in Vietnam, regionally, in ASEAN?. And then, of course, you're based out of Germany, globally. What are some of the trends, perspectives you're seeing?

Dietmar - Tyme: 39:17

If I focus specifically on on, on Vietnam. The one thing that excites me about Vietnam is the the approach that the government has taken in that environment. And I think you know, if you were to just look at some of the recent news releases, how they focusing on chip manufacturing and how they've invested and incentivized not just their own investment, but external investment in in that environment, as well as driving a lot of these big players and collaborating with them. I think on the fifth of December, there was a there was a release of a joint partnership between in the Vietnamese government and the video, where they not just setting up a AI data center, but also AI research facility in the Philippines, which is in the Philippines and Vietnam, which is a joint venture between the the Vietnamese government. And the video, so that government being so proactive and pushing that, I think, is extremely positive. And then the other thing that they've not done, which you which you do see in other geographies, is that they've become too overbearing too quickly with regulation, where they try to regulate, AI too quickly, where, yes, obviously you need to have regulation. You can't just have people do whatever they want without without any guidance, but also you don't want to be too over. You don't want to push it too far, too quickly, to stifle innovation. So that's the thing in sort of the Asian countries, that really excites me about how they will embrace and honestly, have the opportunity to leapfrog other geographies that have taken different, different approaches. You know, there's this really good education systems in in the Asian countries, there's high digital adoption, together with how the governments have embraced, you know, I've mentioned the Vietnamese government, but they're not the only one. And just how it's being embraced, I think it's going to, going to set up those populations in in really good stead to not just to keep up with where, where the world is moving, but to actually take advantage and and and ultimately become leaders, leaders in that environment, yeah, because this, this ultimately is, is, is very much still a place of of innovation, where people need to experiment and understand and they need to be be given that that freedom to do that, obviously you need to still protect your population. And make sure that you still have the right guardrails, but then also just embrace that the technology is there for the people to do it, that you've got the data centers, you've got the research facilities. So, so just looking at that in in those markets, for me, is very, very exciting.

Vanessa Kwan: 42:18

Understand, I think, not only AI developments, but like you rightly pointed out, AI governance, AI regulations, is also certainly something that we see are lacking in this part of the region, but certainly something that we are all looking forward to very much. Ditma Bo, thank you very much for sharing your valuable insights and experiences with us today. I do hope that the audience here do take away a lot of key pointers that you shared with them in terms of how you've structured the data and AI teams, and of course, having the different data people on site to gather insights from the different teams in the respective countries that you operate in. Thank you again for your time, and we hope to continue the conversations with you soon.

Dietmar - Tyme: 42:58

Yeah, thank you for having me. Had a great time. Thank you.

Unknown: 43:04

We hope you've enjoyed the episode. For more information about business growth in the ASEAN region, please visit our website, www, dot IoT business, hyphen platform.com, you.